Pages: 27-34
Introduction. Paid advertising platforms have changed a great deal once machine learning moved into the bidding stack, and Google Ads is now the obvious case in point. I look at how AI driven Smart Shopping campaigns compare with manually managed Search campaigns from a Romanian online wine retailer, and I ask, in particular, what last click attribution misses when the two types of campaigns run in parallel. The dataset combines three Google Analytics reports covering six concurrently running campaigns, plus an assisted conversion analysis on a 30 day lookback window. Four indicators carry most of the weight: CTR, CPC, ecommerce conversion rate and ROAS. What I find is uneven. Smart Shopping pulls in most of the clicks and most of the assisted touches, yet records the lowest last click conversion rate; manual Search campaigns, targeting users already close to buying, score the opposite way. Within this Google Ads account, from my digital marketing agency (Performance Target), approximately 47.6% of all conversions are identified as assisted conversions rather than last click interactions. From a strategic standpoint, this substantial share indicates that relying solely on last click reporting provides an incomplete picture, making it an unreliable foundation for budget allocation decisions.
Aim of the study. To address these complexities, this paper concludes with practical insights from my agency, for small e commerce retailers on how to effectively balance AI driven and manual campaigns. Specifically, it outlines strategies to avoid over investing in branded search while ensuring that top of funnel campaigns which actively generate the initial customer demand are not inadvertently starved of budget.
